Neuro-Symbolic AI: How AWS runs smarter and more secure systems

To make artificial intelligence smarter, safer and more consistent with human reasoning, Neuro-Symbolic AI blends data-driven learning with symbolic logic. This hybrid approach combines the pattern recognition power of neural networks with the structured inference of symbolic systems.

Recognizing this possibility, Amazon Web Services Inc. is investing in neurotumor AI by providing infrastructure, tools and research to scale hybrid systems. The tech giant combines deep learning advances with formal logic to unlock new frontiers with AI reasoning. Byron Cook (Photo), a vice president and well-known scientist at AWS.

Byron Cook from AWS talks with TheCube about what's available for neurosymbolic AI.

“Now, with investments in generator and agent AI, we're feeling comfortable again,” Cook said. “These areas are blurred by what we call neuroreactive AI, but they are a very hot and huge opportunity for us. There are products announced, what's going forward, and a few others where science continues under the hood that drives innovation.”

Cook spoke to TheCube's John Fourier in TheCube + NYSE Wired: AI + Cloud Leader Media Week EventLiveStreaming Studio on Siliconangle Media during an exclusive broadcast on TheCube. They discussed how AWS accelerates with neurotumor AI.

Neuro-Drives deeper into the AI of immunology

Neural networks are great at learning from raw data, while symbolic AI thrives with logic and inference, but wanes with unstructured inputs. By combining both intensities, neurosymbolic AI allows a system that can pave the way for more intelligent and adaptive AI, Cook noted.

“For example, you can train to integrate more data using it. You can combine it with augmented learning, and then you can bring out sidecar tools,” he said. “After the time of inference or inference, statements coming from the linguistic model can be moved, put them in logic, and then prove or disprove their accuracy. Automated inference is the symbolic manipulation of moving symbols and inferring that they are true about the semantics they formally represent.

Cloud computing is now being enhanced Generation AI Adoption by streamlining development, deployment and scalability. According to Cook, this shift is turning Gen AI into a global utility and driving innovation across the industry.

“Gen AI is a paradigm shift from the way ML researchers worked previously, as they were able to build these huge models. “In my field, it's the same. We were wearing shoes on a sequential microprocessor running these algorithms on a laptop or machine under a desk. But because we have these distributed systems, we can store a lot of data, so it really changes the game about how to infer.”

Cloud computing is the democratization of AI by creating powerful and scalable computing that can be accessed by more users. AWS is leading this price with tools IAM Access Analyzeraccording to Cook.

“Now, the cloud has made the algorithms more distributed. It's really an explosion,” he said. “Tools are practical right now and you can put them together. Customers really needed these tools themselves before they deployed, which led to IAM Access Analyzer and VPC Reachability Analyzer. We are currently using the exact same tools we currently use to address hallucination inaccuracies, using automated inference checks and bedrock guardrails.”

Here's a complete video interview, part of the Silicon Angle and part of TheCube coverage TheCube + NYSE Wired: AI + Cloud Leader Media Week Event:

https://www.youtube.com/watch?v=32HFGU16668

Photo: Silicon Angle